scholarly journals A machine learning based delta-radiomics process for early prediction of treatment response of pancreatic cancer

2019 ◽  
Vol 3 (1) ◽  
Author(s):  
Haidy Nasief ◽  
Cheng Zheng ◽  
Diane Schott ◽  
William Hall ◽  
Susan Tsai ◽  
...  
Keyword(s):  
Machine Learning ◽  
Pancreatic Cancer ◽  
Treatment Response ◽  
Prediction Models ◽  
Feature Space ◽  
Response Prediction ◽  
Pathological Response ◽  
Early Prediction ◽  
Linear Regression Models ◽  
Ct Guided

Abstract Changes of radiomic features over time in longitudinal images, delta radiomics, can potentially be used as a biomarker to predict treatment response. This study aims to develop a delta-radiomic process based on machine learning by (1) acquiring and registering longitudinal images, (2) segmenting and populating regions of interest (ROIs), (3) extracting radiomic features and calculating their changes (delta-radiomic features, DRFs), (4) reducing feature space and determining candidate DRFs showing treatment-induced changes, and (5) creating outcome prediction models using machine learning. This process was demonstrated by retrospectively analyzing daily non-contrast CTs acquired during routine CT-guided-chemoradiation therapy for 90 pancreatic cancer patients. A total of 2520 CT sets (28-daily-fractions-per-patient) along with their pathological response were analyzed. Over 1300 radiomic features were extracted from the segmented ROIs. Highly correlated DRFs were ruled out using Spearman correlations. Correlation between the selected DRFs and pathological response was established using linear-regression-models. T test and linear-mixed-effects-models were used to determine which DRFs changed significantly compared with first fraction. A Bayesian-regularization-neural-network was used to build a response prediction model. The model was trained using 50 patients and leave-one-out-cross-validation. Performance was judged using the area-under-ROC-curve. External independent validation was done using data from the remaining 40 patients. The results show that 13 DRFs passed the tests and demonstrated significant changes following 2–4 weeks of treatment. The best performing combination differentiating good versus bad responders (CV-AUC = 0.94) was obtained using normalized-entropy-to-standard-deviation-difference-(NESTD), kurtosis, and coarseness. With further studies using larger data sets, delta radiomics may develop into a biomarker for early prediction of treatment response.

scholarly journals Pancreatic Cancer Early Detection Using Twin Support Vector Machine Based on Kernel

Symmetry ◽  
2020 ◽  
Vol 12 (4) ◽  
pp. 667
Author(s):  
Wismaji Sadewo ◽  
Zuherman Rustam ◽  
Hamidah Hamidah ◽  
Alifah Roudhoh Chusmarsyah
Keyword(s):  
Machine Learning ◽  
Pancreatic Cancer ◽  
Support Vector Machine ◽  
Early Detection ◽  
Feature Space ◽  
Kernel Functions ◽  
Twin Support Vector Machine ◽  
Polynomial Kernel ◽  
Support Vector ◽  
Rbf Kernel

Early detection of pancreatic cancer is difficult, and thus many cases of pancreatic cancer are diagnosed late. When pancreatic cancer is detected, the cancer is usually well developed. Machine learning is an approach that is part of artificial intelligence and can detect pancreatic cancer early. This paper proposes a machine learning approach with the twin support vector machine (TWSVM) method as a new approach to detecting pancreatic cancer early. TWSVM aims to find two symmetry planes such that each plane has a distance close to one data class and as far as possible from another data class. TWSVM is fast in building a model and has good generalizations. However, TWSVM requires kernel functions to operate in the feature space. The kernel functions commonly used are the linear kernel, polynomial kernel, and radial basis function (RBF) kernel. This paper uses the TWSVM method with these kernels and compares the best kernel for use by TWSVM to detect pancreatic cancer early. In this paper, the TWSVM model with each kernel is evaluated using a 10-fold cross validation. The results obtained are that TWSVM based on the kernel is able to detect pancreatic cancer with good performance. However, the best kernel obtained is the RBF kernel, which produces an accuracy of 98%, a sensitivity of 97%, a specificity of 100%, and a running time of around 1.3408 s.

scholarly journals Feasibility of diffusion-weighted magnetic resonance imaging in evaluation of early therapeutic response after CT-guided microwave ablation of inoperable lung neoplasms

2021 ◽  
Author(s):  
Thomas J. Vogl ◽  
Emad H. Emara ◽  
Elsayed Elhawash ◽  
Nagy N. N. Naguib ◽  
Mona O. Aboelezz ◽  
...  
Keyword(s):  
Treatment Response ◽  
Lung Neoplasms ◽  
Microwave Ablation ◽  
Residual Activity ◽  
Early Prediction ◽  
Value Assessment ◽  
Adc Value ◽  
Ct Guided ◽  
Significant Difference

Abstract Objective To determine the early treatment response after microwave ablation (MWA) of inoperable lung neoplasms using the apparent diffusion coefficient (ADC) value calculated 24 h after the ablation. Materials and methods This retrospective study included 47 patients with 68 lung lesions, who underwent percutaneous MWA from January 2008 to December 2017. Evaluation of the lesions was done using MRI including DWI sequence with ADC value calculation pre-ablation and 24 h post-ablation. DWI-MR was performed with b values (50, 400, 800 mm2/s). The post-ablation follow-up was performed using chest CT and/or MRI within 24 h following the procedure; after 3, 6, 9, and 12 months; and every 6 months onwards to determine the local tumor response. The post-ablation ADC value changes were compared to the end response of the lesions. Results Forty-seven patients (mean age: 63.8 ± 14.2 years, 25 women) with 68 lesions having a mean tumor size of 1.5 ± 0.9 cm (range: 0.7–5 cm) were evaluated. Sixty-one lesions (89.7%) showed a complete treatment response, and the remaining 7 lesions (10.3%) showed a local progression (residual activity). There was a statistically significant difference regarding the ADC value measured 24 h after the ablation between the responding (1.7 ± 0.3 × 10−3 mm2/s) and non-responding groups (1.4 ± 0.3 × 10−3 mm2/s) with significantly higher values in the responding group (p = 0.001). A suggested ADC cut-off value of 1.42 could be used as a reference point for the post-ablation response prediction (sensitivity: 66.67%, specificity: 84.21%, PPV: 66.7%, and NPV: 84.2%). No significant difference was reported regarding the ADC value performed before the ablation as a factor for the prognosis of treatment response (p = 0.86). Conclusion ADC value assessment following ablation may allow the early prediction of treatment efficacy after MWA of inoperable lung neoplasms. Key Points • ADC value calculated 24 h post-treatment may allow the early prediction of MWA efficacy as a treatment of pulmonary tumors and can be used in the early immediate post-ablation imaging follow-up. • The pre-treatment ADC value of lung neoplasms is not different between the responding and non-responding tumors.

scholarly journals Treatment Response Prediction in Hepatitis C Patients using Machine Learning Techniques

2021 ◽  
Vol 1 (2) ◽  
pp. 79-89
Author(s):  
Ashfaq Ali Kashif ◽  
Birra Bakhtawar ◽  
Asma Akhtar ◽  
Samia Akhtar ◽  
Nauman Aziz ◽  
...  
Keyword(s):  
Machine Learning ◽  
Hepatitis C ◽  
Treatment Response ◽  
Nearest Neighbor ◽  
Response Prediction ◽  
Machine Learning Techniques ◽  
Support Vector ◽  
K Nearest Neighbor ◽  
Learning Techniques ◽  
Treatment Response Prediction

The proper prognosis of treatment response is crucial in any medical therapy to reduce the effects of the disease and of the medication as well. The mortality rate due to hepatitis c virus (HCV) is high in Pakistan as well as all over the world. During the treatment of any disease, prediction of treatment response against any particular medicine is difficult. This paper focuses on predicting the treatment response of a drug: “L-ornithine L-Aspartate (LOLA)” in hepatitis c patients. We have used various machine learning techniques for the prediction of treatment response, including: “K Nearest Neighbor, kStar, Naive Bayes, Random Forest, Radial Basis Function, PART, Decision Tree, OneR, Support Vector Machine and Multi-Layer Perceptron”. Performance measures used to analyze the performance of used machine learning techniques include, “Accuracy, Recall, Precision, and F-Measure”.

scholarly journals Enhancing Machine Learning Models With Prior Physical Knowledge to Aid in VIV Response Prediction

2021 ◽  
Author(s):  
Leixin Ma ◽  
Themistocles L. Resvanis ◽  
J. Kim Vandiver
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Reynolds Number ◽  
Prediction Error ◽  
Prediction Models ◽  
Response Prediction ◽  
Vibration Response ◽  
Response Amplitude ◽  
Physical Knowledge ◽  
Input Parameters

Abstract Practical engineering prediction models for flow-induced vibration are needed in the design of structures in the ocean. Research has shown that structural vibration response may be influenced by a large number of physical input parameters, such as damping and Reynolds number. Practical response prediction tools used in design are inevitably a compromise between complexity and simplicity of use. Modern machine learning tools may be used to identify which input parameters are most important. Standard machine learning techniques enable the researcher to compile a list of the most important input parameters, ranked or ordered by the effect of each on the prediction error of the model. When all inputs are treated as equals, blind application of machine learning may lead to predictions that are inconsistent with prior physical knowledge. To address this problem, we conducted a parameter selection process using a prior knowledge-based, trend-informed neural network architecture. This approach was used to identify features important to the prediction of the cross-flow vibration response amplitude of long flexible cylinders, given the known prior effect of Reynolds number and damping. The model balances the usual goal of minimizing the model prediction error, but doing so in a manner that closely follows the extensive knowledge we have of the influence of Reynolds number and damping parameter on response. The resulting neural network model was able to reveal additional insights, including the role of mode number shifting, mode dominance and travelling waves in the regulation of VIV response amplitude.

Sign in / Sign up

Export Citation Format

Share Document